Too complex to manage, too big to ignore

A small purchase has one advantage: it can slip under the radar. A few thousand small purchases have exactly the same advantage, but not remotely the same cost. We asked 88 procurement leaders why this part of the spend remains so hard to manage, what it really takes up, and how to start getting a grip on it.

Ask a procurement team whether scattered small purchases deserve better management, and you'll struggle to find anyone who argues otherwise.

Now ask who looks after them, with what resources and what targets for the year. The conversation gets rather more interesting.

In our 2026 Tail Spend Study, 51% of the procurement leaders surveyed consider the subject important, but haven't made it a formal priority. Only 28% rank it among their top five strategic priorities for 2026.

That gap speaks volumes. The problem is well known. But to put resources behind it, you need to be able to explain what it costs, what there is to gain and who is going to do the work. With tail spend, those three answers rarely sit in the same place.

"Small purchase" doesn't mean "small workload"

Tail spend refers to the mass of low-value or infrequent purchases that sit on the fringes of the big, structured categories. It is often likened to class C purchasing, although scope and thresholds vary from one company to the next.

It might be a maintenance call-out, a training course, a rush print job, a software subscription or a one-off piece of consultancy. Different needs, raised by different people, with suppliers whose only common ground is sometimes an entry in your accounts.

From a procurement director's chair, the trade-offs make sense. You start with the multi-million contracts, the critical suppliers and the negotiations whose results you can demonstrate. The €1,200 request can wait. Then another one arrives. And another.

Each order stays modest. The work needed to find a supplier, onboard them, get the purchase approved and process the invoice doesn't necessarily shrink in proportion. The quote is smaller. The form usually has just as many boxes.

Laurence Laroche, a former procurement director at La Poste, Rockwool and Verallia among others, pointed out in our interviews that this focus on the big-ticket items naturally leaves tail spend to one side, even though its cumulative cost and the risks attached can be significant.

For me, this is where the difficulty starts. We prioritise by how much is being bought, when part of the problem lies in the work it takes to buy it.

Everyone handles a piece of it. Who's looking at the whole?

In 28% of responses to our survey, each category buyer manages their own share of tail spend. In 22%, nobody is clearly in charge. So one response in two describes ownership that is either fragmented or missing altogether.

Fragmentation isn't necessarily a problem in itself. A specialist buyer may well be best placed to handle the small requests in their category. But someone has to be able to see what's happening across the business: how many suppliers are in play, which channels are being used and which tasks keep repeating.

Pascal Pelon, who spent more than fifteen years in procurement leadership at AXA, summed up the paradox in our interviews: because tail spend is everywhere, it ends up being nobody's responsibility.

Yet Finance has a direct interest in getting involved. Setting up supplier records, processing invoices and making payments tie up its teams too. Even so, only 19% of respondents say their finance function treats the issue as a priority shared with Procurement.

Appointing an owner is a start. Giving them access to the data, allies in the business and the authority to change how purchasing works is the essential next step. Otherwise, all you've really done is hand them a problem that everyone else will now be delighted to pass their way.

The cost is real. It just doesn't have its own budget line.

Asked whether they had ever quantified the cost of handling their small suppliers, 38% of procurement leaders said no. Only 27% report having carried out a dedicated internal study.

Hard, then, to make the case for investment. You know how much is being bought, but far less about the time spent making those purchases possible: gathering information, checking documents, setting up accounts, resolving discrepancies and chasing people.

Respondents who gave an estimate put the median cost of onboarding a supplier at €600, followed by €500 a year to manage them. These figures rest on 43 and 36 responses respectively. They are benchmarks drawn from the survey, to be tested against each organisation's own processes.

By way of illustration, applying the €500-a-year assumption to 2,000 suppliers gives €1m in management costs. At 5,000 suppliers, the figure reaches €2.5m. This isn't an extra invoice landing on anyone's desk: it's an estimate of the resources tied up in keeping things running.

And you need to be careful about what you then call a saving. Cutting admin can free up time without immediately reducing staff costs. That time has value, but Finance will be right to ask what becomes of it. A shorter backlog? More cases handled? Extra capacity for strategic purchasing?

The first job, then, is to measure a handful of real transactions. How many people are involved in setting up a supplier? How long do they spend on it? Which requests need several rounds of back-and-forth? That gives you a far more useful baseline than a theoretical cost applied uniformly across the entire supplier base.

Visibility doesn't start with a dashboard

Lack of visibility and control tops the list of challenges cited in the study, at 26%, ahead of compliance risk and process costs, both on 21%.

You might conclude that what's missing is a reporting tool. But before you can display the spend, you have to be able to recognise it.

The same service can be recorded under several descriptions, paid by different entities and filed under different categories. Conversely, a single supplier may sell several types of service, so its name alone doesn't always tell you what was bought.

To get a first usable view, you need to bring purchases together around shared data points: supplier, type of service, amount, entity, requester and channel used. Even on a limited scope, this work already lets you tell recurring needs apart from genuinely one-off purchases.

That distinction matters. If ten teams regularly buy the same service, there may be scope to consolidate or to set up a shared preferred supplier. If the ten purchases are all different and one-off, the priority is more likely to be simplifying how they are processed.

Skip this work and you can still produce a very handsome chart. It will simply tell you, in more colours, that you're still short of information.

One channel for everything? Good luck with that.

Tail spend covers needs that are far too varied to push them all down the same route. Buying a monitor, calling out a technician and selecting a consultant don't require the same information or the same level of expertise.

Our survey reflects that variety. Respondents could select more than one channel:

Running several channels side by side can make perfect sense. The trouble starts when nobody knows which one to use, or when the information collected no longer allows purchases to be tracked together.

A clearer set-up starts with a rule users can understand: for this type of need, here's the route, the approvals required and who to contact when something falls outside it.

A catalogue may suit standardised products. A supplier panel can make certain expertise easier to access. Specialist handling may be the answer for a more complex category. The routes differ, but the rules on accountability and the tracking data need to stay consistent.

And you have to look at what the person doing the buying actually goes through. If the official route takes three days for a need they can sort out in twenty minutes, they'll find another way. Before treating that as a discipline problem, it's worth understanding what the process really asks of them.

Why not just hire someone?

Lack of internal resources is the most common reason given for tail spend being neglected, cited by 39% of respondents. But 24% also consider the return on investment too low compared with other priorities.

The two go hand in hand. The issue takes work, while the benefits are hard to prove before you start.

Hiring can strengthen oversight. It doesn't automatically make handling every small request by hand a viable proposition. And your new recruit won't become an expert in industrial maintenance, legal services and marketing procurement between two meetings.

So you have to decide what warrants specialist attention and what can be standardised. A recurring, well-defined request doesn't need to be rebuilt from scratch every time. An unusual, sensitive or technically complex service deserves closer attention, even when the amount is small.

Financial thresholds remain useful. They work better alongside the nature of the need and the risk involved. A small purchase might, for instance, give a supplier access to company data or involve work on a sensitive site.

AI gets interesting once you know what you're asking of it

In our survey, 49% of procurement leaders are exploring the use of AI on tail spend without a concrete plan. 18% have a pilot under way and 6% say they already use it in production.

The interest is there. Execution is lagging some way behind.

To make a useful start, you need to pin down a specific task: classifying spend lines, extracting the details from a request, flagging missing documents or preparing a comparison of bids. You can then measure the time saved, the errors and the checks still needed.

AI also needs rules. What information is mandatory? Which criteria are used to compare two proposals? When should a buyer be brought in? If nobody knows who signs off a purchase, reading the documents faster won't settle the argument.

Olivier Villard, CPO at Mexens, stressed in our interviews the productivity gains that AI brings. For tail spend, that's the crux of it: reducing the work each request requires lets the team take on purchases it previously had to leave to one side.

That's where I see its value. You can extend the service you provide without the admin growing at the same rate. Experts keep their time for assessing a proposal, handling an exception or negotiating what's worth negotiating.

Choose a starting point where you can prove something

In the study, Meetings & Events, professional services and maintenance each receive 23% of first-place votes among the categories contributing most to tail spend. Marketing and communications follow on 18%, then IT on 12%.

These results offer pointers, but the right starting point depends on your organisation. You need to weigh the size of the spend against the realistic scope for improving it.

I'd start with a category where you can get at the data, bring in a business owner and test a change on a defined scope. Picking the biggest line item isn't much use if nobody can help you understand what's in it.

The aim of a first project is to produce results robust enough to decide what comes next. That means a measured baseline, a clearly defined way of working and users who are genuinely involved.

Events as a worked example

At Naboo, we obviously have a view on this category. Above all, it offers a concrete example of what a build-up of small requests can amount to.

In the model built from our client data, €1m of Meetings & Events spend equates to 218 events, with a median value of €1,400. With direct booking, the model assumes around 174 new supplier accounts are set up each year.

Combining process costs avoided and procurement levers, the business case estimates gross annual value of around €220,000, assuming full centralisation. Roughly half comes from procurement levers, the other half mainly from admin avoided.

This result remains a model, and depends on the assumptions used. But it shows why it pays to look at the price of a service and the cost of buying it together. For the practical side, from budgets to deposits and invoices, we go into the subject in our article on managing event spend.

Before you sign up for savings, agree on how they're calculated

Performance-based fees can make it easier to get a project off the ground. But you still need to define what counts as a result.

Is a price reduction measured against the first quote, the last purchase or a comparable offer? Is the scope the same? Are the partner's fees deducted? How much spend actually goes through the new channel?

At the very least, I'd separate three types of value: savings on purchases, external costs avoided and internal time freed up. All three can justify the project, but they don't show up in the same way, either in the accounts or in the teams' day-to-day work.

For a pilot, a handful of indicators is enough to ground the discussion:

• The share of in-scope purchases that actually go through the intended channel.
• The time taken to handle a request, including the back-and-forth.
• The number of supplier set-ups avoided.
• Price differences on comparable services, net of fees.
• Exceptions, and the difficulties users run into.

Adoption matters as much as theoretical potential. A set-up that performs brilliantly on paper but handles a small fraction of purchases will deliver a small fraction of the expected results.

These observations are what allow you to widen the scope, adjust the journey or move on to another category. And you'll have something more useful than agreement in principle: proof that a different way of handling these purchases works.

Almost everyone already agreed that tail spend deserves attention. The next meeting should be about what has actually changed.

About the study

Naboo's Tail Spend Study 2026 is based on a survey conducted in June and July 2026 among 88 procurement leaders from 75 organisations across 12 sectors, supplemented by interviews with Laurence Laroche, Pascal Pelon and Olivier Villard.

Response bases vary by question: 84 respondents for purchasing channels, 82 for the category ranking, 43 for the cost of onboarding a supplier and 36 for its annual management cost. Totals for multiple-choice questions may exceed 100%.

The events business case uses aggregated, anonymised Naboo client data covering April 2025 to September 2026. It presents modelled gross value, before deducting the cost of the solution and the internal effort required to roll it out.

Read the full study: Tail Spend: too complex to manage, too big to ignore

Sources

1. "Tail spend": CIPS, Don't manage tail spend, eliminate it, a report produced with GEP. It argues that procurement has traditionally judged the effort needed to reduce tail spend not worth it, and makes the case for rethinking procurement capability. cips-download.cips.org
2. "class C purchasing": Manutan, glossary entry Tail spend: definition. It describes tail spend as small purchase amounts that are often wrongly treated as unimportant or non-critical. manutan.com
3. "compliance risk": Ministry of Justice, Bribery Act 2010 guidance, on GOV.UK. It sets out six principles for "adequate procedures", including risk assessment and due diligence. gov.uk
4. "company data": Information Commissioner's Office, Contracts and liabilities between controllers and processors. It explains what a contract with a processor must contain and the responsibilities and liabilities of each party under the UK GDPR. ico.org.uk
5. "The interest is there": Deloitte, 2025 Global Chief Procurement Officer Survey. This twelfth edition draws on more than 250 CPOs across 40 countries. deloitte.com
6. "managing event spend": Naboo, page on estimate spend and savings.